How bitcoin miners are powering the AI boom

Yahoo

2 days ago

Some bitcoin (BTC-USD) miners are pivoting to artificial intelligence (AI) as the rapidly evolving tech's data center and power demands expand.
BitFarms (BITF) CEO Ben Gagnon joins Catalysts with Madison Mills and Interactive Brokers chief strategist Steve Sosnick to dive into the phenomenon.
To watch more expert insights and analysis on the latest market action, check out more Catalysts here.
Well, high performance computing plays a critical role in the world of artificial intelligence, providing the computational power needed to train, deploy, and scale intelligence systems quick and efficiently.
The video's results bring into focus the power demands needed to support AI.
Joining us to discuss about Ben Gannon.
He's Bitfarm's CEO.
Bitfarms is a Bitcoin mining company that is expanding existing data centre infrastructure to support HPC and AI.
Workloads to meet the growing energy demands of the industry.
Ben, it's always great to speak with you.
We were just talking on set about how we kind of need to see the proof in the pudding when it comes to AIROI, and you are a great person to talk to about this.
How and where are you seeing the investment in AI playing out in terms of high performance computing and infrastructure going forward?
It's a great question.
You know, for a company like us who've really been focused on building data centres for the last few years that were powering Bitcoin, uh, I think the important thing to focus on is what does these uh new business fras mean for your multiple and your evaluation?
And as a Bitcoin miner, most companies in our space usually trade around a 3 to 5 times multiple, but data centre companies trade around a 20 to a 30 times multiple.
So because of those contracted revenues, because of that certainty over the long period of time, you're able to unlock a lot of value for shareholders and so on, I think for the investors who are looking around at HBC and AI and trying to figure out where are the opportunities after Nvidia has had such a huge run up over the last 10 plus years.
You know, where else are you going to get those those opportunities, and the reality is that as Navidia continues to sell more GPUs and as these GPUs consume more and more power, it's really the infrastructure which is going to be the bottleneck and the really key element here that's going to enable all these companies to benefit from AI.
And to the question of which companies benefit from AI one thing we've been circling this morning is the idea of digital assets benefiting and cryptocurrency players benefiting as well.
How are you seeing that relationship developing going forward?
Oh, it's completely changed the industry almost overnight, you know, 12 months ago there were very few companies that were looking away from pivoting from Bitcoin mining and crypto into HBC and AI, but the math is really, really compelling.
And when you look around at the data centre companies that exist today, there's kind of two buckets.
You have the traditional data centre companies who have really been focused on building large portfolios of small sites, and then you've got the Bitcoin miners who were focused on building small portfolios of large sites.
And when you look at what the data centre needs are for not only for today but for the future, you know, the reality is there's no data centre that's built today that's going to power the GPUs of tomorrow.
And so the scale that Bitcoin miners have gives a very, very fast pathway and a very more cost effective approach towards building that new infrastructure a lot faster.
And Ben, my co-host for the hour, Steve Salznick has a question for you as well.
Hi Ben.
The, the, you know, the thing that strikes me here is Nvidia, call it 34 years ago, was really the all-in company during the first sort of, I'm sorry, Bitcoin run in the first crypto run.
It's almost like the, the industry's following Nvidia just a little bit late.
Am I wrong in thinking that's, you know, that that that analogy is, is, is correct or incorrect, that it's sort of, you know, Nvidia, Nvidia sets the path and, and, and the, and the people who use their chips, uh, into extensively follow along.
Well, Nvidia is, is really setting the pace here in terms of the technological development of these chips, um, but you know, computing is something that's growing at an exponential rate for the last 80 years since it was invented, you know, with Alan Turing and the Enigma machine in World War II.
Um, you know, with Bitcoin mining, you know, we've had the very, very fast growth rate over the last 16 years and so there's a lot of things that are taking place right now where, um, the math is just really, really changing at such a point.
Such a fast pace that you need to be agile, you need to be able to adapt.
You need to be able to change your approach to the changing market demands because, you know, HPC and AI before Chat GPT was really for most people, a theoretical thing, um, but when Chat GPT came out and you had a consumer facing application, it became very real and very tangible for people, and that seemed to spark a gold rush, but, uh, I think from where I'm standing and from the customers' conversations that we're having.
You know, the demand is is just infinite, and we really have barely even begun to scratch the surface of what's possible here with these new chips and this new kind of computing.
And I hear you that you're seeing that demand be infinite, but one of the key questions for investors is whether that demand is going to be infinite for Nvidia.
If we do see soft earnings results from Nvidia this evening, how does that impact your business and your confidence in that demand picture for AI?
Well, the bottleneck on growth for Compute has always been power outside of the silicone shortage of 2021.
So chips are available, they're always going to be available, and Nvidia develops at such a fast pace that you know there's a new generation of chips roughly every 12 months.
But at the back end you always need to have the power to plug in those chips and operate them so that you can actually.
value out of them and you can generate that return on your investment in the GPUs.
So from our side, we haven't seen any reduction in demand even when the Deepse news came out a couple of months ago, there was no reduction in demand.
There was no slowing down of these infrastructure buildouts because they're something that may be taking 2 to 5 years to build out one of these large data centre campuses, and you really can't get distracted by a headline that's going to put you back.
Ben, really appreciate you joining us and giving us that context.
Thank you so much.
Thanks for having me.